Site Audit and AI crawlers
Whether the crawlers behind ChatGPT, Gemini, Claude and Perplexity are allowed to read your site, and what your homepage tells them.
Updated
In short. Site Audit reads your
robots.txtand reports whether five AI crawlers are allowed in. It also shows what a crawler can extract from your homepage. A blocked crawler cannot learn about you from your own site.
The crawlers Peach checks
| Crawler name in robots.txt | Who runs it |
|---|---|
GPTBot |
OpenAI, for ChatGPT |
Google-Extended |
Google, for Gemini and AI features |
ClaudeBot |
Anthropic, for Claude |
anthropic-ai |
Anthropic, an older name |
PerplexityBot |
Perplexity |
For each one, the audit says allowed or blocked. When a crawler is blocked, it shows the rule in your robots.txt that blocks it, so you know which line to change.
How to read the result
- All allowed. Nothing to do here.
- One or more blocked. Decide whether that is deliberate. Some sites block AI crawlers on purpose. If you want AI answers to draw on your site, a block works against you.
- No robots.txt. Crawlers treat a missing file as permission to read everything.
A rule under User-agent: * applies to every crawler that has no group of its own, including these.
Allowing a crawler
To allow a crawler explicitly, give it its own group:
User-agent: GPTBot
Allow: /
Repeat the group for each crawler you want to allow.
What AI understands about your site
The audit also shows the information an AI crawler can extract from your homepage. If your homepage shows almost nothing without JavaScript, a crawler sees almost nothing too.
This matters twice over. It affects what crawlers learn, and it affects Peach: the classification behind your prompts and competitors starts from the same page.
What the audit does not do
- It checks permission, not behaviour. Being allowed in does not mean a crawler has visited.
- It reads
robots.txtonly. It does not checknoindextags, login walls or firewalls. - Allowing crawlers does not raise your mention rate by itself. It removes a barrier. What gets you named is content that answers the question.